memory -go
A Model Context Protocol server that provides knowledge graph management capabilities.
claude mcp add --transport stdio okooo5km-memory-mcp-server-go memory-mcp-server-go
How to use
The Memory MCP Server for Go provides a persistent knowledge graph where you can create, read, update, and delete entities and their relationships. It stores data in SQLite with automatic JSONL migration and supports a flexible transport layer (stdio, SSE with keep-alive, and streamable HTTP). Clients can perform memory-centric operations like adding entities with custom types, linking them via relations, and annotating them with observations over time. Useful tools exposed by the server include creating multiple entities, creating relations between entities, adding or deleting observations, deleting entities or relations, and querying the graph through read/search endpoints. Typical workflows involve initializing a set of entities, establishing relationships to form a graph, and then querying or updating the graph as conversations progress, ensuring memory is preserved across sessions.
To interact with the server, run the binary and connect via the available transport modes. For example, you can use standard input/output (stdio) for local testing, or set up Server-Sent Events (SSE) with a keep-alive mechanism for long-running sessions. The server supports robust error handling and optional client-driven generation sampling, making it suitable for integration with LLM workflows that need a consistent memory store.
How to install
Prerequisites:
- A supported OS (Linux, macOS, or Windows) with a compatible shell.
- curl and tar for download and extraction (for binary install).
- Optional: a PATH directory (e.g., ~/.local/bin) to place the memory-mcp-server-go executable.
Option 1: Quick Install (macOS/Linux)
- Run the installer script to fetch the latest release and install to your PATH:
curl -fsSL https://raw.githubusercontent.com/okooo5km/memory-mcp-server-go/main/scripts/install.sh | bash
Optional arguments:
- -s --version: specify a version, e.g. -s -- -v v0.2.3
- -d --dir: install directory, e.g. -d /usr/local/bin
2) Verify the installation and version:
```bash
memory-mcp-server-go --version
Option 2: Download Pre-built Binary
- Download the appropriate binary release for your platform from the GitHub Releases page (examples below):
- macOS ARM64: memory-mcp-server-go-darwin-arm64.tgz
- macOS x86_64: memory-mcp-server-go-darwin-amd64.tgz
- Linux x86_64: memory-mcp-server-go-linux-amd64.tgz
- Linux ARM64: memory-mcp-server-go-linux-arm64.tgz
- Windows x86_64: memory-mcp-server-go-windows-amd64.zip
- Extract and move the binary to a location in your PATH
# example for macOS/Linux
BIN=$(tar -tzf memory-mcp-server-go-*.tgz | head -1)
tar -xzf memory-mcp-server-go-*.tgz
chmod +x "$BIN"
mkdir -p ~/.local/bin
mv "$BIN" ~/.local/bin/memory-mcp-server-go
- Ensure the location is in your PATH and verify:
memory-mcp-server-go --version
Windows users should extract the ZIP and place memory-mcp-server-go.exe in a PATH directory.
Additional notes
Tips and considerations:
- The server uses SQLite with automatic JSONL migration for compatibility and performance. If upgrading, existing data will migrate to SQLite seamlessly.
- Transports supported: stdio, SSE (with keep-alive), and Streamable HTTP. Configure your client accordingly to maintain connections.
- Memory graph operations include create_entities, create_relations, add_observations, delete_entities, delete_observations, delete_relations, read_graph, search_nodes, and open_nodes. Each operation has specific payload shapes described in the README.
- If you experience permission or PATH issues on macOS/Linux, ensure the target install directory is included in PATH and has execute permissions.
- Use the --version command to confirm you’re running the expected binary release.
Related MCP Servers
trpc-agent-go
trpc-agent-go is a powerful Go framework for building intelligent agent systems using large language models (LLMs) and tools.
station
Station is our open-source runtime that lets teams deploy agents on their own infrastructure with full control.
tiger-cli
Tiger CLI is the command-line interface for Tiger Cloud. It includes an MCP server for helping coding agents write production-level Postgres code.
gopls
MCP server for golang projects development: Expand AI Code Agent ability boundary to have a semantic understanding and determinisic information for golang projects.
kubernetes
A Model Context Protocol (MCP) server for the Kubernetes API.
gcp-cost
💰 An MCP server that enables AI assistants to estimate Google Cloud costs, powered by Cloud Billing Catalog API and built with Genkit for Go